Though I do sometimes use these services, since a few years I block these spiders on my own sites and those of clients.
Often this would happen: Client or user finds an image through Google Image Search on another blog or website, with unclear copyright. Then the client or user would upload this to the server. Then the client, or me, would receive a letter from a (GettyImages) lawyer: If we would please pay for the full licensing right of that 160x160 pixel image.
Thinking about it, I find accommodating to these services can lead to nothing but trouble: Either legal trouble, or hit-and-run users stealing your images, because you paid for higher resolution.
I hope I can separately block this Google service from Google Image Search. Although Google Image Search isn't as good to webmasters as it used to be (especially for those that rely on advertisement clicks) and the users it can send can be negligible: Ranking higher in Google Image Search seems to correlate to ranking higher in Google Web Search.
If it is part and parcel of Google Image Search, I might reconsider my robots.txt directive for Googlebot-Image. They just now opened this up for the public, but it is likely they are already using this internally to gauge (media) quality factors on-page.
P.S.: It would be interesting to see what happens when Google doesn't partner up with GettyImages or iStockPhoto, like in the early days of TinEye you could abuse that service to find the same stock images without watermarks, on the sites of people that already paid for that image.
P.P.S.: Now you can add RDFa or Microdata to mark up your images with a copyright statement, what would happen to sites that host copyrighted images, tagged "not for reproduction"? Google should be able to find the canonical image and "punish" those that don't comply with its copyright.